1. Volume 63, Issue 5, Pages 852-864 (September 2016)
Cas3-Derived Target DNA Degradation Fragments Fuel Primed CRISPR Adaptation 
Tim Künne, Sebastian N. Kieper, Jasper W. Bannenberg, Anne I.M. Vogel, Willem R. Miellet, Misha Klein, Martin Depken, Maria Suarez- Diez, Stan J.J. Brouns 
Molecular Cell 
Volume 63, Issue 5, Pages (September 2016)
DOI: /j.molcel
Copyright © 2016 Elsevier Inc. Terms and Conditions

2. Molecular Cell 2016 63, 852-864DOI: (10.1016/j.molcel.2016.07.011)
Copyright © 2016 Elsevier Inc. Terms and Conditions

3. Figure 1 Plasmid Loss and Transformation Assay
Plasmid loss was assessed by plating cells and scoring for the GFP signal at various time points after induction of cas genes. Individual assays can be seen in Figure S2. The bona fide target is abbreviated as WT.
(A) Example curves and CRISPR PCR of four different types of plasmid behaviors that were observed: rapid plasmid loss without spacer integration (D+P−), delayed plasmid loss and spacer integration (D+P+), strongly delayed plasmid loss and spacer integration (D−P+), and no plasmid loss with no spacer integration (D−P−).
(B) Summary of plasmid behavior of all mutants, showing timing of first plasmid loss and time of first observable spacer integration.
(C) The relative transformation efficiency is plotted for all mutant plasmids (fold change compared with co-transformed non-target plasmid, log2 scale). Bars are color coded on the basis of plasmid behavior classification. Error bars represent SEM of triplicate experiments. The positions of mutations are indicated schematically for each mutant (position 1, bottom; position 32, top). Open ovals represent mutations on positions 6, 12, 18, 24, and 30. Closed ovals represent mutations outside of those positions (effective mutations). The amount of effective mutations is indicated above or below the schematic.
For a more detailed overview of the mutations, see Figure S1.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions

4. Figure 2 Electrophoretic Mobility Shift Assay and Cas3 Activity Assay
All mutants are classified according to previously identified plasmid behavior. The mean and SD for each group are indicated.
(A) Electrophoretic mobility shift assay (EMSA) of the mutant plasmid set. The affinity ratio (amplitude/Kd) is plotted for each mutant (see Table S3 for more details). The bona fide target is abbreviated as WT.
(B) Cas3 DNA degradation activity assay of mutant plasmid set. The initial Cas3 DNA cleavage rate (percentage per minute) is plotted for each mutant.
Individual gels for all activity assays can be found in Figure S4.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions

5. Figure 3
Next-Generation Sequencing Analysis of Cas3 DNA Degradation Products
(A) Left: schematic of R-loop formed by binding of Cascade to dsDNA target. Right: schematic showing the four distinct Cas3 cleavage sites in dsDNA target.
(B) Length distribution of Cas3 DNA degradation fragments of M4 target.
(C) Heatmap of nucleotide frequencies around cleavage sites. The cleavage site is between positions −1 and 1. Positions indicated in black are on the fragments; positions indicated in gray are outside of fragments.
(D) Heatmap of dinucleotide frequencies around cleavage sites. Abundance of dinucleotides was measured in a shifting frame within four nucleotides around the cleavage sites.
See also Figure S6.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions

6. Figure 4 In Vitro Spacer Acquisition Assays
(A) Illustration of the three types of assays performed. In the oligo assay, pCRISPR is incubated with Cas1-2 and a spacer oligo (BG7415/6), leading to half-site integration. In assay 1, pTarget and pCRISPR are incubated with Cascade, Cas3, and Cas1-2 for simultaneous degradation of pTarget and half-site integration into pCRISPR. In assay 2, pTarget is incubated with Cascade and Cas3, and the resulting DNA degradation products are then separately incubated with pCRISPR and Cas1-2.
(B) Gel electrophoresis of integration assay 1. The bona fide target is abbreviated as WT. Left gel, untreated; right gel, proteinase K treated. Cas1-2 presence causes upward shift of DNA. Original plasmids are supercoiled (SC); half-site integration causes nicking of pCRISPR, resulting in the open circular conformation (OC).
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions

7. Figure 5 Half-Site Integration PCR
(A) Illustration of the half-site integration PCR. Primer sets are chosen to show integration into site 1 (leader-proximal repeat end) and site 2 (leader distal repeat end) and to see both possible orientations of the integrated spacer. Primer sequences were chosen on the basis of frequently incorporated spacers (hot spots) in vivo (Fineran et al., 2014).
(B) Gel electrophoresis of half-site integration PCR on the basis of integration assay 2 (left) and oligo assay (right). PCR products representing integrations are indicated with an arrow. PCR products were specific to reactions containing all components. Lower running PCR products are primer dimers (verified by sequencing).
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions

8. Figure 6 Sequencing Analysis of Spacer Integration
(A) Frequencies of exact integration locations for integration at site 1 (gray bars) and site 2 (black bars) as determined by sequencing. The x axis gives the backbone nucleotide to which the spacer is coupled. Frequencies of coupled spacer nucleotides are indicated for the two canonical insertion locations.
(B) Top: schematic of integrated fragment and method of length determination. Bottom: length of the integration amplicon for site 1 and site 2.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions

9. Figure 7 Model of Primed Spacer Acquisition
Cleavage of a targeted plasmid during direct interference by Cascade and Cas3. Cleavage products are near spacer length and reanneal to form duplexes with 5′ and/or 3′ overhangs. The fragments are enriched for NTT sequences on their 3′ ends. A fraction of the duplexes fulfils spacer precursor requirements: 3′ overhangs, CTT at one 3′ end, and a 33 nt distance between the C and the opposite 3′ overhang. Cas1-2 binds spacer precursors with a preference for ideal duplexes as described above (Nuñez et al., 2015a; Wang et al., 2015). The precursor is processed by Cas1-2 to a length of 33 nt with 3′ cytosine. In parallel to processing, 3′ ends of the precursor perform a Cas1-2 catalyzed nucleophilic attack on the two integration sites of the repeat (Nuñez et al., 2015b; Rollie et al., 2015). Integration at the leader-repeat junction occurs first (Nuñez et al., 2016); subsequently, the PAM-derived 3′ cytosine is integrated to ensure correct orientation and production of a functional spacer. A stable spacer integration intermediate is formed (Arslan et al., 2014). The gaps are filled in and repaired by the endogenous DNA repair systems, including DNA polymerase I (Ivančić-Baće et al., 2015).
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions